TTS training code

[acul3]

hello , i notice in script , you provide pretrain for speech encoder and video (CMIIW), but i cant seem to find code for training tts
(vita/model/vita_tts)
AFAIK, vita 1.5 support end-to-end tts

is there a code for training this (vita/model/vita_tts) ?

[BradyFU]

Hi, thanks for your attention on our work! As for the TTS part, you can refer to https://github.com/VITA-MLLM/Freeze-Omni

[acul3]

Hey @BradyFU
Can you pinpoint where is the training code for tts in freeze-pmni, all i can see script for inference

Thank you